OmniFocus FAQs

Question 1

How does the task completion feature ensure accuracy?

Accepted Answer

The `omnifocus_complete_task` tool prioritizes using a unique task ID for completion. If a task ID is not provided, it can use an exact task name and project match. For safety, it will return an error and complete nothing if no match is found or if multiple tasks match the provided criteria, preventing accidental completions.

Question 2

Does it support filtering and managing tasks across different projects and tags?

Accepted Answer

Absolutely. You can query incomplete tasks by specific project name or tag (partial match). When adding new tasks, you can assign them to a project, and when creating new projects, you can specify their type and assign them to existing folders.

Question 3

Are there any prerequisites to use OmniFocus-MCP?

Accepted Answer

Yes, you need a macOS operating system, OmniFocus 3 or 4 installed, and Node.js 20+ to run the server. It also requires an MCP-compatible client (like Claude Code) and initial macOS permissions for automation control over OmniFocus.

Question 4

What specific actions can my AI assistant perform in OmniFocus with this tool?

Accepted Answer

Your AI assistant can query incomplete tasks with various filters (project, tag, flagged status), add new tasks with details (project, due date, tags), update existing tasks by ID, create new projects (parallel/sequential, with folder assignment), and mark tasks as complete with safety checks.

Question 5

What is OmniFocus-MCP and how does it connect AI to OmniFocus?

Accepted Answer

OmniFocus-MCP is a Model Context Protocol (MCP) server that enables AI assistants to programmatically interact with your OmniFocus tasks and projects on macOS. It acts as a bridge, allowing AI to execute commands like querying, adding, updating, and completing tasks directly within OmniFocus.

OmniFocus

OmniFocus

主な機能

ユースケース

主な機能

ユースケース